Inference of Dynamic Origin-Destination Matrices with Trip and Transfer Status from Individual Smart Card Data

Short abstract The research presents an enhanced approach to infer dynamic and multimodal Origin-Destination (OD) matrices from incomplete smart-card validations in Lisbon public transport. Abstract Urban public transport is essential to meet passengers’ mobility needs and contribute to a more sustainable and multimodal mobility in cities. Societal goals such as mobility decarbonisation, aligned with reaching long-term climate-neutrality targets, and emerging health and social equity issues reinforce the need of transport operators to provide high-quality public transport services to replace individual car-based travel. For this purpose, the provision of seamless and multimodal public transport supply requires a complete understanding of the real traffic dynamics and mobility patterns within the city-functional region. This is often challenged by the full availability of public transport data and other features of individual trip data in each context.The research presented in this paper aims at addressing the above challenges by proposing an approach to infer dynamic and multimodal Origin-Destination (OD) matrices from smart-card validations in public transport. More specifically, this approach enables to:i) detect multimodal commuting patterns from individual trips;ii) efficiently detect vulnerabilities on the network pertaining to time and distance spent on transfers and trips;iii) decompose traffic flows in accordance with calendrical rules and user profiles;Additionally, the research develops unimodal and multimodal models for alighting bus stop inference. This latter model accounts for the fact that automated fare collection (AFC) systems operate with only entry-or-exit control/ card validations in some contexts. For example, this is the case found in the city of public bus transport in Lisbon (entry/boarding card validations with no exit control data).This work is being conducted in the context of the “ILU: Integrative Learning from Urban Data and Situational Context for City Mobility Optimization” (DSAIPA/DS/0111/2018), an innovating and pioneering project in the field of artificial intelligence that is committed to optimizing the urban mobility in the Lisbon city by combining multiple sources of traffic data. The Lisbon city is, in fact, used as the study case in this work, with traffic flow analysis being performed from raw smartcard validations gathered from the primary bus operator, CARRIS, and subway operator, METRO.The research reported contributes with an analysis of multimodal public transport data to study passengers’ flow behaviour and estimate dynamic and multimodal OD matrices. We propose alighting stop inference models over the passengers’ paths in the absence and presence of multimodal views, offering the possibility to parameterize maximum walking distances and waiting times on route transfers, extended classical assumptions, and further statistical/confidence *annotations*In addition, the proposed approach for inferring OD matrices yields four unique contributions. First, we allow inference to consider multimodal commuting patterns, detecting individual trips undertaken along different operators. Second, we support dynamic matrices OD inference along parameterizable time intervals and calendrical rules, and further support the decomposition of traffic flows according to the user profile. Third, we allow parameterization of the desirable spatial granularity and visualization preferences. Fourth, our solution efficiently computes several statistics that support OD analysis, helping with the detection of vulnerabilities throughout the transport network. More specifically, statistical indicators related to travellers’ functional mobility needs (commuters for working purposes, etc.), walking distances and trip durations are supported. The inferred dynamic and multimodal OD matrices are the outcome of a developed software which has graphical facilities and guarantees of usability.Results from the case study using data gathered from the two main public transport operators in the city of Lisbon (Bus and Metro) show that 70% of alighting stops can be estimated with high confidence degree from bus smart-card data. Moreover, adding smart-card data from Metro enables an improvement of 10% in accuracy estimation.Since the analysis of multimodal patterns showed that nearly 20% of the journeys within the Lisbon's transportation network require one or more transfers, the inferred OD matrices allowed the identification of critical stops/stations in the network, offering the Bus Public Operators in Lisbon new knowledge and a means to better understand multimodal dynamics and validate OD assumptions.